Professional Documents
Culture Documents
org
J. Math. Comput. Sci. 11 (2021), No. 6, 7459-7475
https://doi.org/10.28919/jmcs/6615
ISSN: 1927-5307
Copyright © 2021 the author(s). This is an open access article distributed under the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract. In this work, we discuss the existence and uniqueness of fixed points for a self-mapping defined on a
C∗ -algebra valued rectangular quasi-metric space. Our results extend and supplement several recent results in the
literature. Some examples are provided to illustrate our results.
Keywords: fixed point; rectangular quasi-metric spaces; C∗ -algebra.
2010 AMS Subject Classification: 47H10, 54H25.
1. I NTRODUCTION
One of the most famous metrical fixed point theorem is the Banach contraction principle [1],
which is the classical tool for solving several nonlinear problems. Based on the noncomplexity
and the usefulness of this principle, it have many extension and generalization into several
directions [4, 5, 6, 7].
∗ Corresponding author
E-mail address: massithafida@yahoo.fr
Received August 09, 2021
7459
7460 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
In 1930, Wilson [9] introduced the concept of quasi-metric spaces. Using this idea many
researcher presented generalization of the renowned Banach fixed point theorem in the qusi-
metric spaces.
In 2000, Branciari [2] introduced the notion of rectangular metric spaces where the triangle
inequality of metric spaces was replaced by another inequality, so-called rectangular inequality.
In 2014, Ma et al. [8] established the notion of C∗ -algebra valued metric spaces by replacing
the range set R with an unital C∗ -algebra, which is more general class than the class of metric
spaces and utilized the same to prove some fixed point results is such spaces.
The following lemma will be useful in our main results.
2. M AIN RESULT
Definition 2.1. Let X be a non empty set. Suppose the mapping d : X × X → A+ satisfies:
Then (X, A+ , d) is a C∗ -algebra valued rectangular quasi- metric space. However we have the
following:
not a C∗
2) (X, A+ , d) is -algebra
valued
asymmetric
metric
space,
as
1 1 1
1 0 2 0 4 0 4 0
= d 2, 1 + d 1
d (2, 0) = = +
4 4,0 .
0 1 0 21 0 14 0 14
3) (X, d) is not aC∗ -algebra
valued
rectangular
metric space, as
1
1 n 0 1 0
= d 2, 1 , for all n ≥ 1.
6=
d n,2 =
1 n
0 n 0 1
7462 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
Definition 2.4. Let (X, A, d) is a C∗ -algebra valued rectangular quasi-metric space and
{xn }n∈N be a sequence in X. Then
(i) We say that {xn }n∈N forward converges to x with respect to A if and only if for given
ε 0A , there is N such that for all n ≥ N, d (x, xn ) ε. We denote it by
lim d (x, xn ) .
n→+∞
(ii) We say that {xn }n∈N backward converges to x with respect to A if and only if for given
ε 0A , there is N such that for all n ≥ N, d (xn , x) ε. We denote it by
lim d (xn , x) = 0A .
n→+∞
lim d (xn , xm ) = 0A .
n,m→+∞
lim d (xm , xn ) = 0A .
n,m→+∞
Remark 2.5. [4] Let (X, d) be as in Example 2.3, { 1n }n∈N∗ be a sequence in X. However we
have the following:
1 1
= 1 and lim d 0, n1 = 1, lim d 2, 1n = 0A . Then,
i) lim d
n→+∞ n,0 = 0A , lim d
n→+∞ n,2 n→+∞ n→+∞
the sequence { 1n } forward converges to 2 and backward converges to 0, so limit is not
unique.
1 1 1 1
ii) lim d
n→+∞ m, n = lim d
n→+∞ m, n = 1. So, forward (backward) convergence dose not im-
ply forward (backward) Cauchy.
Lemma 2.6. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space and {xn }n be a
forward (or backward) Cauchy sequence with pairwise disjoint elements in X. If {xn }n forward
converges to x ∈ X and backward converges to y ∈ X, then x = y.
Proof. Let ε 0A . First assume that {xn }n is a forward Cauchy sequence, so there exists n0 ∈ N
ε
such that d(xn , xm ) 3 for all m ≥ n ≥ n0 . Since {xn }n forward converges to x so there exists
C∗ -ALGEBRA VALUED RECTANGULAR QUASI-METRIC SPACES 7463
ε
n1 ∈ N such that d(xn , x) 3 for all n ≥ n1 . Also {xn }n forward converges to y so there exists
ε
n2 ∈ N such thatd(y, xn ) 3s for all n ≥ n2 . Then for all N ≥ max{n0 , n1 , n2 },
ε ε ε
d(x, y) d(x, xn ) + d(xn , xn+1 ) + d(xn+1 , y) + + = ε.
3 3 3
As ε 0A was arbitrary, we deduce that d(x, y) = 0A , which implies x = y. When {xn }n is a
backward Cauchy sequence, the proof is similar to an earlier state
Definition 2.7. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space. X is said
to be forward (backward) complete if every forward (backward) Cauchy sequence {xn }n in X
forward (backward) converges to x ∈ X.
Definition 2.8. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space. X is said to
be complete if X is forward and backward complete.
Definition 2.9. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space. A mapping
T : X → X is a C∗ -valued contractive mapping on X, if there exists an a ∈ A with k A k< 1 such
that
for all x, y ∈ A.
Theorem 2.10. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space and let T :
X → X is a C∗ -valued contractive mapping on X, then there exists a unique fixed point in X.
Proof. It is clear that if A = 0A , then T maps the X into a single point. Thus without loss of
generality, one can suppose that A 6= 0A . Choose an x0 ∈ X and set xn+1 = T xn = T n x0 . Notice
that in C∗ -algebra, if a, b ∈ A+ and a b, then for any x ∈ A both x∗ ax and x∗ bx are positive
elements and x∗ ax x∗ bx.
Substituting x = xn−1 and y = xn , from (1), for all n ∈ N, we have
a∗ d (xn−1 , xn ) a
...
(a∗ )n d (x0 , x1 ) an
a∗ d (xn−1 , xn+1 ) a
(a∗ )2 d (xn−2 , xn ) a2
...
(a∗ )n d (x0 , x2 ) an
Case 1: Assume that m = 2l + 1 with l ≥ 1. By property (ii) of the C∗ -algebra valued rectan-
gular quasi-metric space, we have
d (xn , xn+1 ) + d (xn+1 , xn+2 ) + d (xn+2 , xn+3 ) + d (xn+3 , xn+4 ) + d (xn+4 , xn+n+2l+1 )
(a∗ )n d (x0 , x1 ) an + (a∗ )n+1 d (x0 , x1 ) an+1 + ... + (a∗ )n+2l d (x0 , x1 ) an2l+1
i=n+2l
= ∑ (a∗ )k d (x0 , x1 ) ak
i=n
i=n+2l 1 1
= ∑ (ak d (x0 , x1 ) 2 )∗ d (x0 , x1 ) 2 ak
i=n
i=n+2l 1
= ∑ |d (x0 , x1 ) 2 ak |2
i=n
i=n+2l 1
k ∑ |d (x0 , x1 ) 2 ak |2 kIA
i=n
1
i=n+2l
kd (x0 , x1 ) 2 k2 ∑ kak2k IA
i=n
1 kak2n
kd (x0 , x1 ) 2 k2 IA → 0A (n → ∞).
1 − kak2
C∗ -ALGEBRA VALUED RECTANGULAR QUASI-METRIC SPACES 7465
d (xn , xn+2 ) + d (xn+2 , xn+3 ) + d (xn+3 , xn+4 ) + d (xn+4 , xn+5 ) + d (xn+5 , xn+n+2k )
(a∗ )n d (x0 , x2 ) an + (a∗ )n+2 d (x0 , x1 ) an+2 + ... + (a∗ )n+2k−1 d (x0 , x1 ) an+2k−1
i=n+2k−1
∗ n
= (a ) d (x0 , x2 ) a + n
∑ (a∗ )i d (x0 , x1 ) ai
i=n+2
1 1
i=n+2k−1 1 1
= (an d (x0 , x2 ) 2 )∗ d (x0 , x2 ) 2 an + ∑ (ak d (x0 , x1 ) 2 )∗ d (x0 , x1 ) 2 ak
i=n+2
1
i=n+2k−1 1
= |d (x0 , x2 ) 2 an |2 + ∑ |d (x0 , x1 ) 2 ai |2
i=n+2
1
i=n+2k−1 1
k|d (x0 , x2 ) 2 an |2 kIA + k ∑ |d (x0 , x1 ) 2 ai |2 kIA
i=n
1 1
i=n+2k−1
kd (x0 , x2 ) 2 k2 kak2n IA + kd (x0 , x1 ) 2 k2 ∑ kak2i IA
i=n
1 1 kak2n
kd (x0 , x2 ) 2 k2 kak2n IA + kd (x0 , x1 ) 2 k2 IA → 0A (n → ∞).
1 − kak2
Therefore xn is a forward Cauchy sequence with respect to A. By the completeness of (X, A, d)
there exists an z ∈ X such that
d (xn+1 , xn ) = d (T xn , T xn−1 )
a∗ d (xn , xn−1 ) a
...
(a∗ )n d (x1 , x0 ) an
7466 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
(a∗ )n d (x2 , x0 ) an
Case 1: Assume that m = 2l + 1 with l ≥ 1. By property (ii) of the C∗ -algebra valued rectan-
gular quasi-metric space, we have
d (xn+m , xn ) = d (xn+2l+1 , xn )
1
i=n+2l
kd (x1 , x0 ) 2 k2 ∑ kak2k IA
i=n
1 kak2n
kd (x1 , x0 ) 2 k2 IA → 0A (n → ∞).
1 − kak2
d (xn+m , xn ) = d (xn+2k , xn )
1
i=n+2k−1 1
|d (x2 , x0 ) 2 an |2 + ∑ |d (x1 , x0 ) 2 ai |2
i=n+2
1
i=n+2k−1 1
k|d (x2 , x0 ) 2 an |2 kIA + k ∑ |d (x1 , x0 ) 2 ai |2 kIA
i=n
1 1
i=n+2k−1
kd (x2 , x0 ) 2 k2 kak2n IA + kd (x0 , x1 ) 2 k2 ∑ kak2i IA
i=n
1 1 kak2n
kd (x2 , x0 ) 2 k2 kak2n IA + kd (x1 , x0 ) 2 k2 IA → 0A (n → ∞).
1 − kak2
lim d(u, xn ) = 0A .
n→∞
C∗ -ALGEBRA VALUED RECTANGULAR QUASI-METRIC SPACES 7467
Since
0A d (T z, z) d (T z, T xn ) + d (T xn , xn ) + d (xn , z)
Hence d (z, u) = θ and z = u, which implies that the fixed point is unique.
Example 2.11. Let A = M2 (R) of all 2 × 2 matrices with the usual addition ,scalar multiplica-
tion and multiplication.
a1 a2 b1 b2
Define partial ordering on A as ⇔ ai ≥ bi for i = 1, 2, 3, 4. For
a3 a4 b3 b4
12
a1 a2
i=4
any A ∈ A we define its norm as, k k = ∑ |ai |2 .
a3 a4 i=1
d(x, y) = 0 ⇔ y = x f or all x, y ∈ X.
7468 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
and
0.3 0
1 1 1
d , = d 0, =
3 4 2 0 0.3
0.2 0
1 1 1
d ,0 = d , =
3 4 2
0 0.2
1 0.35 0
1 1
d 0, =d , =
3 2 4 0
0.35
0.6 0
1 1 1 1
, =d , =
d
3 2 3 2 0 0.6
|x − y| 0
d (x, y) = otherwise.
0 |x − y|
Then (X, A+ , d) is a C∗ -algebra valued rectangular quasi-metric space. However we have the
following:
Indeed
1 1
x4 − y4 0
d(T x, Ty) = 1 1
0 x −y
4 4
√1 0 x−y 0 √1 0
3 3
0 √1 0 x−y 0 √1
3 3
= a∗ d(x, y)a.
where
√1 0
a= 3
0 √1
3
with verify
√
2
kak = √ ≤ 1.
3
case 2 : x ∈ [1, 2], y ∈ A. Then
1
1
T (x) = x 4 , T (y) = 1, D (T x, Ty) = 2 x 4 − 1 .
1
x4 − 1 0
d (T x, Ty) = 1
.
0 x −1
4
Indeed
1
x −1
4 0
d(T x, Ty) = 1
0 x −1
4
√1 0 x−1 0 √1 0
3 3
0 √1 0 x−1 0 √1
3 3
√1 0 x−y 0 √1 0
3 3
0 √1 0 x−y 0 √1
3 3
= a∗ d(x, y)a.
where
√1 0
a= 3
0 √1
3
with verify √
2
kak = √ ≤ 1.
3
Hence, the condition (1) is satisfied. Therefore, T has a unique fixed point z = 1.
Example 2.12. Let A = M2 (R+ ) of all 2 × 2 matrices with the usual addition
,scalar
multi-
a1 a2 b1 b2
plication and multiplication. Define partial ordering on A as
a3 a4 b3 b4
12
a1 a2
i=4
⇔ ai ≥ bi for i = 1, 2, 3, 4. For any A ∈ A we define its norm as ,k k = ∑ |ai |2 .
a3 a4 i=1
it follows that
Indeed
1 x y 1
0 e −e 0 0
2 2
d (T x, Ty) if x ≥ y
1 1
0 2 0 0 0 2
1 1
0 0 0 0
d (T x, Ty) 2 2
if x ≤ y
0 12 0 e−x − e−y 0 12
where
1
2 0
a=
1
0 2
with verify
1
kak = < 1.
2
Hence, the condition (1) is satisfied. Therefore, T has a unique fixed point z = 0.
Definition 2.13. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space. A mapping
0
T : X → X is a C∗ -valued Kannan-type mapping on X, if there exists an a ∈ A+ with k a k< 1
2
such that
Theorem 2.14. Let (X, A, d) a C∗ -algebra valued rectangular quasi-metric space and let T :
X → X is a C∗ -valued Kannan-type mapping on X, then there exists a unique fixed point in X.
7472 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
0
Proof. Without loss of generality, one can suppose that a 6= 0A . Notice that a ∈ A+ ,
a [d(x, T x) + d(y, Ty)] is also a positive element. Choose an x0 ∈ X and set xn+1 = T xn = T n x0 .
Substituting x = xn−1 and y = xn , from (6), for all n ∈ N, we have
0
Since a ∈ A+ with k a k< 12 , using Lemma 1.1, I − a is invertible and also k(I − a)ak < 1. Thus
...
(I − a)−n an [d (x0 , x1 )]
= hn [d (x0 , x1 )]
with hn = (I − a)−n an .
Substituting x = xn−1 and y = xn+1 , from (1), for all n ∈ N, we have
(I − a)−2 a2 [d (xn−2 , xn )]
...
(I − a)−n an [d (x0 , x2 )]
= hn [d (x0 , x2 )]
Case 1: Assume that m = 2l + 1 with l ≥ 1. By property (ii) of the C∗ -algebra valued rectan-
gular quasi-metric space, we have
d (xn , xn+1 ) + d (xn+1 , xn+2 ) + d (xn+2 , xn+3 ) + d (xn+3 , xn+4 ) + d (xn+4 , xn+n+2l+1 )
1
i=n+2l k
kd (x0 , x1 ) 2 k2 ∑ kh 2 k2
i=n
1 khk2n
kd (x0 , x1 ) 2 k2 IA → 0A (n → ∞).
1 − khk2
d (xn , xn+2 ) + d (xn+2 , xn+3 ) + d (xn+3 , xn+4 ) + d (xn+4 , xn+5 ) + d (xn+5 , xn+n+2k )
i=n
2n
n 1
2 khk
khk kd (x0 , x2 ) k + kd (x0 , x2 ) k
2 IA → 0A (n → ∞).
1 − khk2
d (xn+1 , xn ) = d (T xn , T xn−1 )
hn [d (x1 , x0 )] .
7474 ABDELKARIM KARI, MOHAMED ROSSAFI AND HAFIDA MASSIT
hn [d (x2 , x0 )] .
This is equivalent to
Since
Hence d (z, u) = θ and z = u, which implies that the fixed point is unique.
C ONFLICT OF I NTERESTS
R EFERENCES
[1] S. Banach, Sur les opérations dans les ensembles abstraits et leur application aux équations intégrales, Fund.
Math., 3 (1922), 133–181.
[2] A. Branciari, A fixed point theorem of Banach-Caccioppoli type on a class of generalized metric spaces, Publ.
Math. Debrecen, 57 (2000), 31–37.
[3] A. Kari, M. Rossafi, E. Marhrani, M. Aamri, θ − φ −contraction on (α, η)−complete rectangular b−metric
spaces, Int. J. Math. Math. Sci. 2020 (2020), Article ID 5689458.
[4] A. Kari, M. Rossafi, H. Saffaj, E. Marhrani, M. Aamri, Fixed-Point Theorems for θ − φ -Contraction in
Generalized Asymmetric Metric Spaces, Int. J. Math. Math. Sci. 2020 (2020), Article ID 8867020.
[5] A. Kari, M. Rossafi, E. Marhrani, M. Aamri. New Fixed Point Theorems for θ − φ −contraction on complete
rectangular b−metric spaces, Abstr. Appl. Anal. 2020 (2020), Article ID 8833214.
[6] A. Kari, M. Rossafi, E. Marhrani, M. Aamri. Fixed-Point Theorem for Nonlinear F-Contraction via w-
Distance, Adv. Math. Phys. 2020 (2020), Article ID 6617517.
[7] M. Rossafi, A. Kari, E. Marhrani, M. Aamri, Fixed Point Theorems for Generalized θ − φ −Expansive Map-
ping in Rectangular Metric Spaces, Abstr. Appl. Anal. 2021 (2021), Article ID 6642723.
[8] Z. Ma, L. Jiang, H. Sun, C∗ −algebra-valued metric spaces and related fixed point theorems, Fixed Point
Theory Appl. 2014 (2014), 206.
[9] W. A. Wilson, On quasi-metric spaces, Amer. J. Math. 53 (1931), 675-684.
[10] G. J. Murphy, C∗ -Algebras and Operator Theory, Academic Press, Inc., Boston, Mass, USA, 1990.